A 96-well plate of crimp-top vials came back from the autosampler with a 4.7% RSD on the internal standard after 40 injections, and the analyst had already thrown out the column. The method was a 60:40 acetonitrile/water gradient at 40 °C, and the septa were a bargain silicone-faced laminate. The silicone had swollen, the PTFE film had wrinkled, and the puncture channel never resealed. Choosing which vial septa material to use for a given solvent and temperature is not a catalog decision. It is a compatibility calculation you either do before the run or you do afterward with a failed sequence.
This guide is about the polymer, not the part number. I am going to walk through PTFE, silicone, EPDM, and the laminates that combine them, and I am going to be specific about where each one works, where it fails, and what the failure looks like on the bench. If you need help with dimensions or fit, the vial septa reference library covers that. Here I am staying on chemistry and temperature.
The first question is not “what temperature?” It is “what is in the vial?” A septum that is perfectly rated for 200 °C will still fail in 30 minutes if it is sitting over dichloromethane, and a septum that shrugs off DCM will deform at 150 °C. You have to run the solvent check first.
Here is the working rule I use. Silicone is a hydrocarbon rubber. It absorbs nonpolar and moderately polar solvents and swells. PTFE is a fluoropolymer with one of the lowest surface energies of any solid, and it is essentially inert to everything you will put in an autosampler vial. EPDM is an ethylene-propylene rubber that handles aqueous, polar, and acidic systems well but has no business near aromatic or chlorinated solvents. The laminate exists because neither pure material does both jobs.
What swelling actually does to a septum
Silicone in toluene or DCM can take up 20 to 40 percent of its weight in solvent. That swelling is not cosmetic. The septum grows in diameter, which changes the compression fit in the cap, and it grows in thickness, which changes the seal geometry. Worse, the absorbed solvent sits in the polymer and slowly diffuses into the headspace. If you are running trace analysis, that absorbed solvent is a continuous background source. I have seen a silicone septum over DCM produce a persistent solvent peak in a GC-MS blank that took three bakeouts to clear.
The pH question is narrower than people think
None of these polymers is a pH buffer, but the ones with fillers and cure residues are. Standard silicone has residual cure byproducts and fumed silica filler that can leach under strongly basic conditions above pH 9 or strongly acidic conditions below pH 2. EPDM handles pH 1 to 12 comfortably because it has a saturated backbone and good chemical resistance to aqueous acids and bases. PTFE is fine across the entire 0 to 14 range. If your method is a strong base digestion or a 5% HCl extraction, the filler and cure system matter as much as the base polymer.
The temperature window for each polymer, and where the window is a lie
Datasheets give you a maximum continuous use temperature. What they do not give you is the temperature at which the septum stops resealing after puncture, which is the number that actually governs an autosampler sequence.
| Material | Continuous use temp | Practical reseal limit | Reseal behavior after puncture | Solvent resistance |
|---|---|---|---|---|
| Pure PTFE (0.010 in / 0.25 mm) | -100 to 260 °C | ~200 °C | Poor alone; cold-flows, does not spring back | Universal, inert to nearly all solvents |
| Silicone (0.040 in / 1.0 mm) | -60 to 200 °C | ~150 °C | Excellent below 150 °C, 50+ punctures typical | Poor in nonpolar/chlorinated; good in polar |
| EPDM (0.040 in / 1.0 mm) | -50 to 150 °C | ~130 °C | Good, 40 to 50 punctures | Good aqueous/polar; poor aromatics |
| PTFE/silicone laminate | -60 to 200 °C | ~150 °C | Excellent, PTFE film plus silicone core | PTFE face inert; silicone core protected |
| PTFE/EPDM laminate | -50 to 150 °C | ~130 °C | Good, suited to aqueous methods | PTFE face plus EPDM core, aqueous/polar |
Read the reseal limit column carefully. A pure PTFE septum rated to 260 °C will hold pressure at 250 °C, but it will not close a 0.5 mm puncture channel after the needle withdraws. PTFE cold-flows under compression. The hole stays a hole. That is why pure PTFE is used on high-temperature headspace and purge-and-trap vials where the needle punctures once or twice, not on a 100-injection liquid autosampler sequence.
The laminate fixes this. A 0.002 in (0.05 mm) PTFE film bonded to a 0.040 in (1.0 mm) silicone core gives you an inert sample-facing surface and a rubber core that springs back. The silicone never touches the sample, so solvent swelling is limited to whatever diffuses through the film. That is the design logic behind most of what you will buy for HPLC and GC autosamplers, and it is why the laminate dominates the HPLC vial market.
Match the polymer to the method, not to the instrument
I have watched people buy septa by autosampler brand for years. The instrument does not care what polymer you load. The method does. Here is how I sort the four families against the methods I actually run.
Reversed-phase HPLC with aqueous/organic mobile phases
PTFE/silicone laminate, 0.040 in thick, is the default. The PTFE face handles the acetonitrile or methanol, the silicone core reseals through 50-plus injections, and the temperature ceiling of 150 °C is well above any column oven you will use. If your method is entirely aqueous (ion chromatography, certain USP assays), a PTFE/EPDM laminate is fine and slightly cheaper. If you are running a autosampler vial at 4 °C in the sample compartment, silicone stays flexible and reseals well at low temperature, which is another point in its favor.
GC with nonpolar solvents
This is where silicone alone gets dangerous. Hexane, toluene, and DCM will swell a bare silicone septum. Use a PTFE/silicone laminate or, for single-puncture headspace, pure PTFE. The PTFE face keeps the solvent off the rubber. If you are doing headspace GC, the vial is pressurized and heated, and the septum sees both temperature and pressure. A laminated septum with a PTFE sample face is the safe choice up to about 150 °C.
Trace analysis and low-level detection
When you are chasing parts per trillion, the septum is a background source. Silicone bleeds cyclic siloxanes (D3 through D10) into the headspace, and those show up as a forest of peaks in GC-MS. If your method is a US EPA SW-846 volatile or semivolatile procedure with low detection limits, look at the SW-846 method requirements and then pick a septum that has been baked and bagged. Pure PTFE has the lowest bleed. A PTFE-faced laminate is second. Bare silicone is the worst offender. I have seen a silicone septum contribute a siloxane background that raised the reporting limit by a factor of three.
High-temperature and aggressive chemistry
Above 150 °C, silicone starts to degrade and lose reseal. Above 200 °C, only PTFE survives, and you accept single-use punctures. For strong acids, strong bases, and oxidizing agents, PTFE is the only safe face. For a method that combines 180 °C with a chlorinated solvent, you are in a narrow corner: pure PTFE, single puncture, and a fresh vial for every injection. That is expensive, but it is the only combination that holds.
Two field cases where the wrong polymer announced itself
Case one: the drifting internal standard
A contract lab running a 60:40 ACN/water method at 40 °C saw the internal standard area fall 12% across a 60-injection sequence, with the RSD climbing to 4.7% by injection 40. The column was fine. The autosampler was fine. The septa were a 0.040 in silicone laminate with a thin PTFE film, and the film had separated from the core at the puncture site after roughly 25 injections. The silicone underneath had swollen from the acetonitrile and was weeping solvent vapor into the headspace, diluting the sample. The fix was a thicker PTFE film (0.002 in instead of 0.001 in) on a silicone core, which held the film in place through 100 injections. The RSD dropped to 0.8%.
The measurement that identified it: a blank vial run at injection 50 showed a solvent ghost peak that had not been there at injection 1. The septum was the source, not the mobile phase.
Case two: the EPDM that dissolved in toluene
An environmental lab running a toluene extraction on a heated shaker switched from silicone to EPDM septa to get better aqueous resistance. Within two hours at 60 °C, the EPDM had absorbed toluene, softened, and extruded through the crimp. The vials leaked, the extract volume dropped, and the recovery fell to 68% against a 85 to 115% acceptance window. EPDM has an aromatic resistance problem, and toluene is a solvent, not a trace contaminant. The fix was a PTFE/EPDM laminate for the aqueous portion of the method and a PTFE/silicone laminate for the toluene portion. The lesson: EPDM is an aqueous polymer. Do not put it over an aromatic solvent just because the temperature rating looks adequate.
Puncture count, torque, and the numbers that decide reseal life
Reseal life is not a single number. It is a function of needle gauge, septum thickness, core hardness, and how much compression the cap puts on the septum. A few working figures:
- Needle gauge: a 22-gauge needle (0.72 mm OD) leaves a smaller channel than a 26-gauge (0.46 mm), but the smaller channel reseals faster. Most autosamplers use 22-gauge, and that is the number I size reseal life against.
- Septa thickness: 0.040 in (1.0 mm) silicone core is the workhorse. Thinner (0.010 in PTFE only) is for single puncture. Thicker than 0.060 in starts to interfere with needle penetration depth.
- Puncture count: a good PTFE/silicone laminate holds 50 to 100 punctures before reseal degrades. Bare silicone is similar. Pure PTFE is 1 to 5.
- Crimp torque: a properly crimped 11 mm aluminum seal on a 20 mm vial should be set so the septum is compressed but not crushed. I use a crimper set to about 1.5 to 2.0 N·m equivalent, and I check with a seal that the septum face is flat, not dimpled. Over-crimping cold-flows the PTFE and shortens reseal life. Under-crimping leaks.
- GPI thread: for screw-top vials, the 9 mm short-thread (GPI 9-425) and 10-425 closures need the septum seated flat before the cap is torqued. Hand-tight plus a quarter turn is the practical target. A torque wrench is overkill, but consistency is not.
If you want the manufacturer’s data on needle compatibility and reseal performance, the Agilent consumables library and the Thermo Fisher sample handling pages publish septum specifications that are worth reading against your method. The Restek technical resources are also good on septa bleed and background for GC.
A selection checklist you can run in five minutes
Before you order, answer these in order. The first “no” sends you to a different polymer.
- Is the sample solvent aromatic, chlorinated, or a strong swelling agent (toluene, DCM, THF, hexane)? If yes, the sample face must be PTFE. Silicone or EPDM alone is out.
- Is the method above 150 °C? If yes, PTFE only, single puncture, fresh vial per injection.
- Is the method aqueous, polar, or acidic/basic between pH 1 and 12? EPDM core is acceptable, preferably as a PTFE/EPDM laminate.
- Is the method a trace GC-MS method with siloxane-sensitive detection? Pure PTFE or a baked PTFE-faced laminate. Avoid bare silicone.
- How many injections per vial? More than 20 punctures means you need a resealing core (silicone or EPDM), not pure PTFE.
- Is the vial stored cold (4 °C) before injection? Silicone stays flexible; EPDM stiffens. Check reseal at the storage temperature, not just at injection.
If you are building a method from scratch, the vial and closure selection should be locked down at the same time as the septum, because the cap compression and the septum thickness interact. A septum that is right for the chemistry but wrong for the cap torque will still leak.
The recommendation, stated as a decision rule
Use a PTFE/silicone laminate (0.002 in PTFE film on a 0.040 in silicone core) as the default for any reversed-phase HPLC or GC method that runs below 150 °C and uses a solvent that would swell bare silicone. Switch to a PTFE/EPDM laminate only when the method is aqueous or polar and the temperature stays below 130 °C. Go to pure PTFE only when the temperature exceeds 200 °C or the chemistry is so aggressive that no rubber can survive, and accept single-puncture use. Never put bare silicone over a chlorinated or aromatic solvent, and never put EPDM over an aromatic one, regardless of what the temperature rating says. The polymer has to match the solvent first and the temperature second, and the failure you avoid by getting that order right is a run you do not have to repeat.
References & further reading
Related resources from our lab consumables network:
- Vial Septa at vialsepta.com
- Hplc Vial at hplcgc.com
- Autosampler Vial at autosamplerhplc.com
- Vial at vials-filters.com
External references and standards:
